Better models, better advice: Why the next generation of stock assessment matters

When discussions turn to sustainable fisheries, attention naturally focuses on quotas, surveys and stock advice. Yet behind every annual catch recommendation lies something few people ever see: a statistical model. These models combine decades of biological observations with fishery-dependent and scientific survey data to estimate how many fish are in the sea, how many young fish are entering the stock, and how much can be harvested sustainably. They are the analytical engines that underpin the scientific advice produced by ICES and, ultimately, the management decisions taken by governments across Europe.

Like every scientific discipline, stock assessment continues to evolve. As new biological understanding emerges and computing power increases, researchers are able to represent the complexity of marine ecosystems with increasing realism. A recent development from DTU Aqua, the new smsR stock assessment framework  discussed by Nils. S Jacobson and colleagues (Ecol Sol and Evidence – 2026 – Jacobsen – smsR A seasonal assessment model for exploited stocks), represents another important step in that progression.

Fish do not live in annual time steps

One of the greatest simplifications in many traditional stock assessment models is that they effectively treat the year as a single unit. This assumption works reasonably well for many long-lived species, but it becomes increasingly problematic for fast-growing forage fish.

Species such as sandeel and sprat can undergo dramatic changes within only a few months. Individuals grow rapidly, mature, spawn, experience changing natural mortality and become available to fisheries at different times of the year. At the same time, scientific surveys are conducted in different seasons, while commercial fishing often occurs during relatively short periods. Ignoring these seasonal dynamics risks creating a mismatch between the biology of the stock and the mathematics used to describe it.

The new smsR framework addresses this challenge by explicitly incorporating seasonal processes into the assessment. Rather than assuming the population remains effectively unchanged throughout the year, the model allows growth, recruitment, natural mortality and fishing mortality to vary between seasons, providing a much closer representation of how fish populations actually behave.

Why seasonality matters

At first glance, incorporating seasonal biology may appear to be a relatively small technical improvement. However, the consequences for stock assessment can be substantial.

Using simulation studies, the authors compared seasonal assessments with more traditional annual models. They found that annual models consistently tended to overestimate both recruitment and spawning stock biomass when seasonal dynamics were important. In practical terms, this means that ignoring seasonal processes can create an overly optimistic picture of stock status.

Reducing this source of bias improves confidence that estimated stock size reflects biological reality rather than artefacts of model structure. For fisheries management, this matters because both overestimation and underestimation carry consequences. Overestimating stock size may result in catch advice that places unnecessary pressure on the resource. Underestimating abundance can lead to overly precautionary catch limits with economic consequences for fishing communities and seafood supply chains. Improving the precision of assessments therefore benefits both conservation and sustainable utilisation.

Particularly valuable for forage fish

The advantages of smsR are especially relevant for small pelagic species, including sandeel, sprat and Norway pout, which are characterised by rapid growth, strong seasonal changes and highly variable recruitment.

These species occupy a central position in marine food webs while simultaneously supporting important commercial fisheries. Their short life spans mean that recruitment success and seasonal growth patterns can have an immediate influence on stock size, making accurate modelling particularly important. The authors therefore recommend the framework specifically for small pelagic stocks, where accounting for seasonal changes in weight, mortality and fishing patterns helps reduce assessment bias.

More than a new model

Although smsR introduces new statistical capabilities, perhaps its greatest contribution is that it provides a modern, transparent and open-source platform for future development.

The software has been written as an R package, making it accessible to researchers while incorporating tools for diagnostics, forecasting, management strategy evaluation and future methodological improvements. It has already been adopted as the operational assessment model for several North Sea pelagic stocks, demonstrating its practical value beyond academic development.  As fisheries science increasingly incorporates environmental drivers, changing natural mortality and ecosystem interactions, flexible modelling frameworks such as smsR provide the foundation on which the next generation of stock assessments can be built.

Science that strengthens sustainable fisheries

The annual advice produced by ICES is only as robust as the evidence and methods that underpin it.

While debates around fisheries management often focus on quotas or policy decisions, advances in assessment methodology are equally important. Every improvement that reduces uncertainty, better reflects fish biology or improves diagnostic performance ultimately strengthens the scientific basis for sustainable fisheries management. For industries that depend on healthy fish stocks—including the marine ingredients sector—continued investment in assessment science is therefore not simply an academic exercise. It is an investment in more reliable advice, more transparent management and, ultimately, healthier marine ecosystems.